Send to

Choose Destination
See comment in PubMed Commons below
Eur J Surg Oncol. 2005 Sep;31(7):707-14.

Continued axillary sampling is unnecessary and provides no further information to sentinel node biopsy in staging breast cancer.

Author information

Academic Surgery and Breast Unit, Royal Marsden NHS Foundation Trust, 203, Fulham Road, London SW3 6JJ, UK.



Sentinel lymph node biopsy (SLNB) has become increasingly accepted as a diagnostic method to stage the axilla in breast cancer, selecting women with a positive sentinel node for completion axillary clearance. As SLNB became established, many surgeons supplemented SLNB to sample a minimum of four lymph nodes, on the assumption that the four-node technique is supported by randomised trial data. We hypothesised that the practice of undirected sampling to supplement SLNB adds little information to the status of the residual axilla.


One hundred and sixty-five patients with early breast cancer were studied. Following successful identification of the sentinel node, 84 women had completion axillary dissection and 81 women had an axillary sample with at least four nodes available for pathological assessment.


Following successful identification of the sentinel node in 165 patients, the false negative rate (FNR) was 2/44=4.5% (95% CI 0.6-15.5), sensitivity 42/44=95.5% (84.5-99.4) and negative predictive value (NPV) 121/123=98.4% (94.2-99.8). In the axillary dissection cohort, the FNR was 2/26=7.7% (0.9-25.1), sensitivity 24/26=92.3% (74.9-99.1) and NPV 58/60=96.7% (88.5-100). In the axillary sample group, the FNR was 0/18=0% (0-18.5), sensitivity 18/18=100% (81.5-100) and NPV 63/63=100% (94.3-100). The SLNB was the only positive node in 12/26 (46.2%) in the axillary dissection group and 10/18 (55.6%) in the axillary sampling group. There was no patient in the axillary sampling group where the sample node was positive and the sentinel node negative.


Once SLNB is validated within the multidisciplinary unit, undirected sampling of the axilla following identification of the sentinel node(s) is unnecessary. The additional sampling of non-sentinel nodes has no role to play either in the assessment of a potential false negative SLNB nor as predictive information on the status of the residual axillary nodes.

[Indexed for MEDLINE]
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for Elsevier Science
    Loading ...
    Support Center